AI for Project Managers

You are an experienced Project Management consultant drafting a Project Charter. This charter is not an academic exercise — it is a decision document that will be presented to the project sponsor. It must be honest, specific, and useful.

Read the project briefing below carefully. Then produce a Project Charter following these expert instructions.

## GENERAL RULES (apply to every section)

1. NEVER SANITIZE. If the project is in trouble, the charter says so. If there are unresolved political conflicts, the charter names them. A charter that pretends everything is fine is worse than no charter at all — it destroys the PM's credibility in the first meeting.

2. EXTRACT, DON'T INVENT. Every claim in the charter must come from the briefing or be a logical inference from it. If the briefing doesn't mention something, don't make it up. If you need information that isn't in the briefing, flag it as "TO BE CONFIRMED" rather than fabricating a plausible-sounding answer.

3. SEPARATE FACTS FROM DECISIONS. A good charter distinguishes between what has been decided, what is assumed, and what still needs a decision. When you find an unresolved issue, mark it explicitly as "DECISION REQUIRED" and state who needs to make that decision and by when.

4. WRITE FOR THE SPONSOR, NOT FOR A TEXTBOOK. The sponsor already knows what a project charter is. They need to understand THIS project's situation in 10 minutes. Cut every sentence that doesn't carry specific, actionable information about this project.

5. THIS IS A CHARTER, NOT A PROJECT PLAN. A charter authorizes the project and frames the key decisions. It does NOT include a WBS, detailed task breakdowns, weekly schedules, resource allocation tables, or procurement plans. If you find yourself writing more than 8 pages, you are writing the wrong document.

---

## SECTION-BY-SECTION INSTRUCTIONS

### 1. PROJECT PURPOSE & JUSTIFICATION

State why this project exists and why it matters NOW. Connect it to:
- The strategic or political context driving urgency
- Any hard deadlines (funding expirations, regulatory dates, political commitments)
- The current state of the project (is this a fresh start or a recovery? has work already begun?)

Common AI mistake to avoid: Writing a generic "this project will improve the community" paragraph. The purpose must explain why the sponsor should care about this specific charter at this specific moment.

### 2. OBJECTIVES & SUCCESS CRITERIA

Define 4-6 measurable objectives. For EACH objective:
- State what will be measured
- State the target value or condition
- State the deadline

Rules:
- At least one objective must address STAKEHOLDER PERCEPTION (community, users, public opinion) — not just delivery metrics. Projects that deliver on time but lose public support are failures.
- At least one objective must address COMPLIANCE or REGULATORY requirements if they exist.
- Success criteria must be specific to this project. "Deliver on time and on budget" is not a success criterion — it's a platitude. What does "on time" mean here? What is the specific budget constraint?

Common AI mistake to avoid: Listing objectives that sound professional but are unmeasurable ("ensure stakeholder alignment", "deliver high-quality results").

### 3. SCOPE STATEMENT

Define three things clearly:

**IN SCOPE:** List the concrete deliverables and work packages.

**OUT OF SCOPE:** List items that someone might reasonably assume are included but are NOT.

**SCOPE UNDER NEGOTIATION:** This is the section most AI-generated charters miss entirely. If there are unresolved scope decisions — competing priorities between stakeholders, budget pressure forcing trade-offs, items that may be cut — list them here honestly. For each:
- State the competing options
- Name who is advocating for each option
- Describe the trade-off (cost, schedule, political impact)
- Recommend a deadline for the decision
- Optionally, state the PM's recommended approach (but frame it as a recommendation, not a decision)

Common AI mistake to avoid: Presenting scope as clean and settled when the briefing clearly shows it isn't. This is the single most damaging error — it sets the PM up to look incompetent when the conflicts surface later.

### 4. KEY STAKEHOLDERS

For EACH stakeholder (person, group, or organization), provide:

| Field | What to write |
|-------|--------------|
| Name / Role | Who they are |
| Interest | What they actually want from this project (their real interest, which may differ from their official position) |
| Influence | High / Medium / Low — and WHY (budget authority? public platform? veto power? political connection?) |
| Potential conflict | Which other stakeholders they are in tension with, and over what |
| Risk if mismanaged | What happens concretely if this stakeholder is ignored or antagonized |

Rules:
- Identify the REAL decision-maker, who may not be the official sponsor.
- If someone has been sidelined or bypassed, note it — they are often the most dangerous stakeholder because they have motivation to undermine.
- Include external stakeholders (community groups, regulators, press) — not just the org chart.

Common AI mistake to avoid: Listing stakeholders as a flat table of Name + Role + "Keep informed." This is useless. The value is in the analysis of interests, conflicts, and consequences.

### 5. KEY RISKS

List the top 5-8 risks, ranked by ACTUAL impact on this project (not by textbook categories).

For each risk:

| Field | What to write |
|-------|--------------|
| Risk | Clear statement of what could go wrong |
| Likelihood | High / Medium / Low — with reasoning based on the briefing |
| Impact | What happens to the project if this materializes (schedule, budget, political, legal) |
| Response strategy | A SPECIFIC action plan, not a generic verb. "Mitigate" is not a strategy. "Schedule weekly meetings with the heritage office to pre-negotiate modifications before formal submission" is a strategy. |
| Owner | Who is responsible for executing the response |
| Deadline | When must the response be in place |

Rules:
- Focus on risks that are SPECIFIC to this project's situation.
- Political risks and stakeholder risks are real risks.
- If the briefing mentions a financial risk, quantify it. Show the numbers.
- If a risk has dependencies (Risk A makes Risk B worse), note the connection.

Common AI mistake to avoid: Listing 10+ generic risks from a PMBOK chapter instead of the 5 specific risks that could actually kill this project.

### 6. HIGH-LEVEL TIMELINE & MILESTONES

Work BACKWARD from the hard deadline. List 6-10 milestones with dates.

Rules:
- Start with the end date and work backward.
- Identify CRITICAL PATH DECISIONS — decisions that are currently blocking progress.
- If an external audit, inspection, or review has a known date, it goes here.
- Be honest about what is realistic.
- This is a charter-level timeline: 6-10 milestones only.

Common AI mistake to avoid: Listing milestones as a neat sequence without acknowledging that some depend on decisions that haven't been made yet.

### 7. BUDGET SUMMARY

Present the budget as a financial snapshot:

- Total budget and funding sources (with conditions or restrictions)
- Amount committed/spent to date
- Contingency reserve — original amount, amount already used, amount remaining
- Known unbudgeted costs
- Financial risk exposure

Use a simple table. Show the math.

Common AI mistake to avoid: Showing budget as a single line without revealing the financial pressure points.

### 8. AUTHORITY & GOVERNANCE

Define:
- Reporting line (address ambiguity or dual-reporting directly)
- Decision authority (what the PM can decide alone vs. escalation)
- Escalation path for each major unresolved decision
- Meeting cadence

Common AI mistake to avoid: Writing a generic governance paragraph.

---

## FORMAT REQUIREMENTS

- Professional, direct language. No filler sentences. No motivational language.
- Use tables for stakeholders, risks, budget, and milestones.
- Mark every unresolved item as ⚠️ DECISION REQUIRED with a recommended deadline.
- Total length: 4-8 pages.

---

## PROJECT BRIEFING

[PASTE YOUR PROJECT BRIEFING HERE]

# Project Briefing: The Meridian Cultural Center

## You are the new Project Manager. Here's what you're walking into.

---

Project: Renovation of the former Meridian textile factory (1920s industrial building, listed as local heritage) into a mixed-use cultural center with co-working spaces, an event hall, and a public library branch.

Location: Meridian District, a historically working-class neighborhood currently undergoing rapid gentrification in a mid-sized European city.

Client: The City Council, through the Urban Renewal Office.

Budget: €4.2M (EU Recovery Fund grant - 70%) + City co-financing (30%). The EU grant expires on December 31, 2026. Any unspent funds are returned. No extensions.

Timeline: Construction must be substantially complete by November 15, 2026, to allow 6 weeks for final inspections and handover before the grant deadline. You are starting now (March 2026). That gives you 8 months.

---

## What you need to know

The previous PM was removed 3 months ago. Officially for "personal reasons." Unofficially, because the project was 2 months behind schedule and the Council's opposition party leaked the delays to the press. The Mayor's office is now directly overseeing the project. Your real boss is the Mayor's Chief of Staff, Ana Delgado, not the Urban Renewal Office director (who resents this).

The architect has a vision. Elena Voss designed the renovation to preserve the factory's industrial character - exposed brick, steel beams, original windows. Beautiful, award-winning design. Problem: the heritage conservation office has rejected two of her proposed structural modifications, and she has refused to redesign them, calling the rejections "bureaucratic ignorance." Elena has a strong public profile and the local press loves her. You can't fire her. You can't ignore the heritage office either.

The neighborhood is divided. Half the residents see the cultural center as a catalyst for community revival. The other half see it as the final step of gentrification - "a co-working space for tech workers, not for us." There is an active neighborhood association, Meridian Roots, that organized protests against the previous version of the project in 2023. They got 2,000 signatures. The protests stopped when the city added the public library branch to the program. If the library gets cut from scope (which the budget pressure makes tempting), expect the protests to restart immediately.

The main contractor is already mobilized, but barely. GreenBuild Corp won the tender 5 months ago and has done demolition and initial structural work. However, they are also working on two other city contracts simultaneously. Their site manager, Tomás Ruiz, is competent but stretched thin. You have heard (not confirmed) that GreenBuild is having cash flow problems and may be delaying material purchases.

There's an asbestos problem nobody budgeted for. During demolition, the previous PM's team found asbestos in the original insulation of the east wing. Removal was not in the original scope or budget. A preliminary estimate puts the cost at €180,000–€240,000. The contingency reserve was €300,000, but €120,000 has already been spent on design changes. You have roughly €180,000 in contingency left - barely enough, and only if the lower estimate holds.

Two Council members want different things. Councilor Martín (Culture) wants the event hall to be the centerpiece - 300-seat capacity, professional acoustics, retractable seating. Councilor Ibarra (Social Services) wants the library to be bigger and argues the co-working space should be reduced. Both report to the Mayor, and the Mayor has not made a decision. You are expected to "find a solution that works for everyone."

One more thing. The EU grant requires that at least 15% of the construction workforce be hired from the local neighborhood as part of the social inclusion clause. GreenBuild has been hiring from their usual subcontractor network. Current local hiring stands at 6%. If this isn't fixed before the mid-project EU audit (scheduled for July), the entire grant could be clawed back.